The approaching exodus of retiring baby boomers will severely erode the knowledge base of many companies. Fortunately, there are ways to re-create this crucial expertise.

The deep smarts of organizations are walking out the door. Deep smarts are the contents in the heads of experts that enable them to make swift, wise decisions based on years of experience. A number of trends threaten to converge into a perfect storm of knowledge loss, especially in the technical- and science-based occupations. While the percentage of highly skilled workers made up of foreign-born scientists and engineers has steadily increased in the past two decades, their native countries now want them back.

And then there is the approaching mass exodus of retirees. In many industries, these employees have knowledge that is essential to organizational success. But financially strapped companies are doing increasingly less on-the-job training, which is essential for transferring these deep smarts to the next generation. In Deep Smarts: How to Cultivate and Transfer Enduring Business Wisdom, Walter Swap and I argue that leaders need to identify, nurture and transfer the know-how comprising the engine of their organizations. Every organization has a few key members whose departure would devastate operations. Of course, not all retirees have a knowledge treasure buried in their craniums. And even those who do are subject to the same social and psychological biases as the general population, biases that can sometimes blind our judgement.

But people with deep smarts can be indispensable. Why? Because their particular brand of expertise is based on long, hard-won experience. They are the go-to people known for their swift, seemingly intuitive judgments. Such experts differ from their less experienced colleagues in having the ability to view a problem at a system level and yet dive into the details when necessary-and identify a familiar pattern.

Consider the example of a rocket scientist who saved the day because of his deep smarts. In the early 1980s, his company was competing with another for a US government contract to produce tactical missiles, worth billions of dollars over the next 30 to 40 years. Each competitor tested six working prototypes; none was satisfactory. The scientist, who was not an official member of the project team, called the primary participants to a meeting. He proceeded to awe the assembly by proposing detailed changes that he had worked out by himself during just one week of concentrated effort. Without notes, for several hours, he walked them through the redesign-from the weapon point to its aft-explaining all the software, wiring and hardware changes that would be required to win the competition. His changes were supported and the company won the contract.

It was rocket science, but the scientist didn't get deep smarts from his formal education alone. His kind of expertise accrued over 20 years of working on all different components of the missiles. In this example, the knowledge that could be lost was not only particular to the industry but also proprietary to the company. When such individuals retire or leave, the company suffers a blow from which it may take years to recover.

Re-creating Deep Smarts

Not all deep smarts are this proprietary, nor are they all technical. Managerial, organizational or interpersonal skills and judgment can also be critical. Whatever the domain of the deep smarts, however, a critical issue for organizational leaders is how such knowledge can be passed along to colleagues with less expertise. Since so much of the expertise is tacit-that is, not articulated or documented-and because it therefore exists mostly in the heads and hands of the experts, it can't be readily shipped from one brain to another.

In fact, Walter and I argue that deep smarts can't really be transferred at all. They must be re-created through the process of guided experience. Based on a three-year study of the transfer of knowledge between coaches and relative novices, we propose at least four kinds of guided experience: practice, observation, problem-solving and experimentation. The emphasis is on guided because all four of these produce deeper smarts in the learner when a knowledge coach interacts extensively with the novice, helping to plan, implement and review the learning experience.

Even the best athletes seek practice under the guidance of an expert coach. But two kinds of observation are also important: shadowing and mind-stretching. Shadowing, as the term suggests, is allowing junior staff to follow more experienced operators around, observing actions, speech, decision making-and then having them discuss their observations with the experts. A more provocative form of observation is to assign learners to a series of unusual experiences designed to challenge assumptions and enlarge worldviews. For example, when Best Buy top management determined that they needed to build an innovative capability in the company, they hired consultancy Strategos. The consultants' task was not the usual "analyze and report", but to coach teams of Best Buy employees through a series of mind-stretching experiences-including travel to destinations as diverse as computer game arcades in Seoul, Korea, and the American Girl store in Chicago (specializing in historical dolls). These direct experiences with foreign environments challenged team members' assumptions about the purpose and design of products, and led to some radically innovative ideas. For example, could Best Buy design a socially desirable place for communal use of computers that would have the powerful allure of a destination (such as American Girl) and (like the game arcades) attract a much younger set of users than usual?

Another mode of re-creating deep smarts is joint problem-solving by expert and protege. Deep smarts are, by their nature, contextual: Experts are often unaware of how they solve certain problems and cannot readily bring up a mental list of all possible strategies and outcomes. Their deep smarts are activated by the situation at hand. By working together, the proteges learn how the experts approach a problem-including using personal networks to narrow the search for information-and the experts often stretch their own mental boundaries.

Finally, guided experimentation is essential when there is no certain knowledge, as is so often the case with swift-changing technology and markets. The Best Buy teams mentioned earlier had to test their ideas out in small market experiments before they could judge their worth. Therefore, they opened a prototype "Bang" computer game arcade to teenagers and carefully observed what ideas worked, from helping customers mix to providing snack foods.

The solution to losing deep smarts, in short, is to invest in their re-creation. If the United States is not to lose its world preeminence, we suggest that we need to take the accelerating loss of such knowledge seriously and begin to staunch the flow.

Dorothy Leonard is the William Abernathy Professor of Business Administration Emerita at Harvard Business School. She can be reached at dleonard@hbs.edu. Please send your comments to Executive Editor Alison Bass at abass@cio.com.